code logs -> 2018 -> Mon, 01 Jan 2018< code.20171231.log - code.20180102.log >
--- Log opened Mon Jan 01 00:00:18 2018
00:22 Kindamoody|afk is now known as Kindamoody
00:51 * McMartin finishes creating the C64 disk image
00:51 * McMartin heads over to the RPi so he can create the final zip file
00:51 * McMartin says incompletely-audible things about the perfidy of filesystems with semantically important metadata tags.
00:52
<&McMartin>
On the plus side, at least Unix zip/unzip understand them.
00:52 Kindamoody is now known as Kindamoody[zZz]
00:54
<@himi>
The C64 filesystem was one such beast, McMartin?
00:55
<&McMartin>
Well, kind of, but it's not the problematic one, because those files are distributed as disk images anyway.
00:55
<&McMartin>
It's the Acorn filesystem that's the issue here.
00:55
<@himi>
Ah
00:55
<@himi>
I forgot your RPi was running RISC OS
00:56
<&McMartin>
(C64 also isn't a problematic one because the only thing it *really* tracks is "executable" and emulators just treat everything not from a disk image as having that bit set)
00:57
<&McMartin>
ADFS has more extensive notions of file type that need to be preserved, but it turns out ZIP has fields precisely for this.
00:57
<&McMartin>
At least they aren't resource forks.
00:58 macdjord is now known as macdjord|out
00:58
<&McMartin>
That RPi also runs Raspbian, but I didn't do anything fun with Raspbian that wouldn't work just as well on any other Linux system on Bumbershoot this year.
01:50 Degi [Degi@Nightstar-2aa9mg.dyn.telefonica.de] has joined #code
02:00 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
02:00 mode/#code [+qo Vornicus Vornicus] by ChanServ
02:05 Kindamoody[zZz] [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer]
02:05 * McMartin finishes up his archiving, and also figures out how to make it work in emulation.
02:20 Vornlicious [Vorn@Nightstar-f868h3.sub-174-210-8.myvzw.com] has joined #code
02:24 Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds]
02:35 Vornlicious [Vorn@Nightstar-f868h3.sub-174-210-8.myvzw.com] has quit [[NS] Quit: Bye]
02:35 Vorntastic [Vorn@Nightstar-60o.3mj.149.47.IP] has joined #code
02:37 Vornlicious [Vorn@Nightstar-f868h3.sub-174-210-8.myvzw.com] has joined #code
02:40 Vorntastic [Vorn@Nightstar-60o.3mj.149.47.IP] has quit [Ping timeout: 121 seconds]
03:18 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
03:18 mode/#code [+o Kindamoody|autojoin] by ChanServ
03:21 Vornlicious [Vorn@Nightstar-f868h3.sub-174-210-8.myvzw.com] has quit [Connection closed]
03:21 Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
03:22 * [R] stares at /var/run/log
03:23
<&[R]>
Whyyyyyyyyyy
03:23
<~Vornicus>
?
03:24
<&[R]>
It's redundant and not even a symlink
03:24
<&[R]>
Also empty
03:28 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed]
03:28
<&[R]>
Given everyone's "let's merge /bin /sbin /usr/bin /usr/sbin" I'm baffled to as why they'd have two locations for logs.
03:30
<@himi>
/var/run is supposed to be a tmpfs mount, isn't it?
03:30
<&[R]>
Is it?
03:30
<&[R]>
It's a symlink to /run here
03:30
<@himi>
Yeah, and /run is definitely a tmpfs mount on this system
03:32
<&McMartin>
2017 Bumbershoot Bullshit compiled and posted. https://bumbershootsoft.wordpress.com/2018/01/01/2017-compilation-and-review/
03:49 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
04:18 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
04:18 mode/#code [+o Kindamoody|autojoin] by ChanServ
04:35 Degi [Degi@Nightstar-2aa9mg.dyn.telefonica.de] has quit [Connection closed]
04:47 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed]
04:49 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
04:49 mode/#code [+o Kindamoody|autojoin] by ChanServ
04:55 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer]
05:06 Vornlicious [Vorn@Nightstar-gsqumb.sub-174-210-12.myvzw.com] has joined #code
05:06 Vornlicious [Vorn@Nightstar-gsqumb.sub-174-210-12.myvzw.com] has quit [[NS] Quit: Bye]
05:06 Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Connection reset by peer]
05:06 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
05:06 mode/#code [+o Kindamoody|autojoin] by ChanServ
05:06 Vorntastic [Vorn@Nightstar-gsqumb.sub-174-210-12.myvzw.com] has joined #code
05:36 Derakon is now known as Derakon[AFK]
05:58 Derakon_ [Derakon@Nightstar-fr5qel.ca.comcast.net] has joined #code
06:01 Derakon[AFK] [Derakon@Nightstar-fr5qel.ca.comcast.net] has quit [Ping timeout: 121 seconds]
07:13 Jessikat` [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
07:13 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Connection closed]
07:29 Jessikat` is now known as Jessikat
07:31 Jessikat` [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
07:31 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Connection closed]
07:31 Vash [Vash@Nightstar-1l3nul.res.rr.com] has joined #code
07:39 Jessikat` is now known as Jessikat
08:00 celticminstrel is now known as celmin|sleep
08:19 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Ping timeout: 121 seconds]
08:22 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
08:22 mode/#code [+o Kindamoody|autojoin] by ChanServ
08:29 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer]
08:33 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
08:33 mode/#code [+o Kindamoody|autojoin] by ChanServ
08:39 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed]
08:55
<&McMartin>
jerith: You need to resubmit your solutions to Opus Magnum: Something about how area is computed has changed.
08:55
<&McMartin>
But even after that, I believe you have some work to do on Voltaic Coil. :)
08:56 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
08:56 mode/#code [+o Kindamoody|autojoin] by ChanServ
08:56
<&jerith>
Thanks for letting me know. :-)
08:57
<&McMartin>
(I believe the new change is that "infinite" products now only charge the area for the amount of product checked, and it used to be much more.)
08:57 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed]
08:57
<&McMartin>
(In particular, Sword Alloy, and Golden/Dark thread were both cases where suddenly I took the lead with no change.)
09:02
<&jerith>
Oh, hey. Tarinaky's showing up in my histograms now.
09:03
<&McMartin>
one of us, one of us
09:03
<&jerith>
... and has beaten us in Surrender Flare.
09:03
<&jerith>
Also a few others.
09:06
<&McMartin>
I've solved a few more production cabinets, and I've got the occasional speed or succinctness advantage on you
09:06
<&McMartin>
You are consistently beating me on cost though.
09:07 * McMartin says a variety of unkind things about Special Amaro.
09:10 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
09:10 mode/#code [+o Kindamoody|autojoin] by ChanServ
09:17 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer]
09:17 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
09:17 mode/#code [+o Kindamoody|autojoin] by ChanServ
09:21 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer]
09:37 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
09:37 mode/#code [+o Kindamoody|autojoin] by ChanServ
09:39 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer]
09:40 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
09:40 mode/#code [+o Kindamoody|autojoin] by ChanServ
09:44 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed]
09:58
<&McMartin>
... a lot of work on Voltaic Coil. :)
09:58
<&McMartin>
I'm beating top percentile on both speed and area now.
09:58 Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
09:58 mode/#code [+o Kindamoody|autojoin] by ChanServ
10:00 Kindamoody|autojoin is now known as Kindamoody
10:07
<&McMartin>
And that's pulling a limiting reagent at the maximum possible speed, so I don't think I'm going to get that one any faster.
10:07 Vash [Vash@Nightstar-1l3nul.res.rr.com] has quit [[NS] Quit: Quit]
10:10
< Vorntastic>
I remember one spacechem level where I pulled one reagent so fast that if it'd asked for more than 40 I'd have crashed it
10:11
< Vorntastic>
Something like 13/2 cycles per output
10:14
<&McMartin>
Opus Magnum has more aggressive limiters, for the most part
10:15
<&McMartin>
Reagents take up space on your grid all the time, and are there for the taking unless you have taken it and haven't cleared it yet.
10:15
<&McMartin>
So the maximum speed you can consume any individual reagent component is once every two cycles.
10:15
<&McMartin>
For this particular one (Voltaic Coil) there is only one metal input and I'm consuming it at max speed.
10:15 Vornlicious [Vorn@Nightstar-lul0ub.sub-174-210-3.myvzw.com] has joined #code
10:15 Vorntastic [Vorn@Nightstar-gsqumb.sub-174-210-12.myvzw.com] has quit [Ping timeout: 121 seconds]
10:16
<&McMartin>
Which in turn means that the steady state cannot possibly be improved. Any gains are going to be in time-to-first-product-created, and apparently I edged a cycle of the top-percentile score there.
10:17
<&McMartin>
But OM solutions tend to be a lot more, hrm, parallel than SpaceChem ones.
10:17
<&McMartin>
It's got a lot more of an assembly line feel to it.
10:17
<&McMartin>
And lots and lots of partial products are continuously in flight, because your waldoes generally are fixed in place and hand bits off to other ones.
10:18
< Vornlicious>
Yeah, spacechem did that for synthesis puzzles and had 1/10 steps in the production ones
10:18
<&McMartin>
I don't understand that sentence
10:19
<&Reiver>
If I tried and bounced off SpaceChem, am I likely to find any satisfaction in Opus Magnum?
10:19
<&McMartin>
I remember SpaceChem would block if you tried to input from an empty pipe or output into a full one
10:19
<&McMartin>
Reiver: That's going to depend a lot on why you bounced off SpaceChem.
10:20
< Vornlicious>
The limiters; you can always use in alpha on synthesis (single reactor) puzzles; on production puzzles the pipes get loaded once every ten cycles
10:20
<&McMartin>
If it was because it's ugly or because Defense missions were beyond your patience, then OM is both pretty and has no time pressure whatsoever
10:21
<&McMartin>
If it is because programming games are homework assignments, well
10:21 Kindamoody [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed]
10:21
<&McMartin>
It hides it a little better than TIS-100 or Shenzhen I/O
10:21 Kindamoody [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code
10:21 mode/#code [+o Kindamoody] by ChanServ
10:22
<&McMartin>
gnolam's objection was that all the puzzles eventually become routing puzzles, but I found it didn't do that until the *very* end and maybe not even until the postgame because your build grid is almost infinite.
10:23
<&McMartin>
OTOH, if "move stuff from a to b" is routing puzzle, then that's everything
10:26
< Vornlicious>
The puzzle in question needed two of a particular reagent for every three outputs, so 19.5 cycles to consume 20 cycles of stuff is pretty tight.
10:26
<&jerith>
Reiver: I bounced off SpaceChem and loved Opus Magnum.
10:27
<&McMartin>
To date SpaceChem is the only Zachtronics game where I've beaten the postgame, but I think OM will be #2 there unless I get thwarted *unusually* hard.
10:28
<&Reiver>
Fair
10:28
<&Reiver>
It is certainly the prettier.
10:28
< Vornlicious>
I still haven't finished spacechem
10:28
<&McMartin>
Also exports as gif instead of .ogv or whatever the fuck it was Spacechem used
10:31
< Vornlicious>
I'm kind of frustrated by spacechem's lack of multiple save states
10:31
<&McMartin>
Oh yeah you get four slots per puzzle in OM
10:31
<&McMartin>
That started at TIS-100 at the latest
10:37
<&jerith>
I have more than four slots for some of my puzzles.
10:38
<&McMartin>
For me it's generally "FUNCTION" "SPEED" "COST" "AREA" and a scratch slot that was probably originally FUNCTION.
10:38
<&jerith>
Just confirmed: six slots used in one of them.
10:38
<&jerith>
I think you get an arbitrary number, like in Shenzhen I/O.
10:40
<&jerith>
Ooh, yeah. I have a *lot* of work to do in Voltaic Coil.
10:41
<&jerith>
I've been bumped off the top in all three sections.
10:42
<&McMartin>
Really, I should be working on the Infinifactory Eye Drops.
10:42
<&McMartin>
But I'm going to bed.
10:42
<&jerith>
I should get back to Infinifactory at some point.
10:42
<&McMartin>
Oh, I mean the, uh, eyedrop of revelation
10:43
<&McMartin>
Because my take on the description is that it made the inventor see Infinifactory in the future.
10:43
<&jerith>
Oh, that one.
10:43
<&jerith>
I read it as "he saw into *our* world".
10:43
<&jerith>
Which is basically Infinifactory.
10:45
<&McMartin>
WIthout the alien overlords, mostly.
10:46
<@TheWatcher>
Probably~
10:47
<&jerith>
Have you *seen* the current crop of world leaders?
10:50
<&McMartin>
Do I look like David Icke to you
11:01
<@Alek>
Our current crop of world leaders certainly is Icke, yes. :P
11:04
<&McMartin>
LIZARD PEOPLE, DEAR READERS
11:17 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds]
11:56
<&jerith>
Two cycles behind.
11:56 * jerith leaves that for now, moves on to other metrics.
12:20 gnolam_ [quassel@Nightstar-o3qe20.priv.bahnhof.se] has joined #code
12:37
<&jerith>
Yay, done.
12:37
<&jerith>
Now I guess I need to optimise some production puzzles.
13:21 gnolam_ [quassel@Nightstar-o3qe20.priv.bahnhof.se] has quit [Connection reset by peer]
13:48 Degi [Degi@Nightstar-2aa9mg.dyn.telefonica.de] has joined #code
13:59 gnolam_ [quassel@Nightstar-o3qe20.priv.bahnhof.se] has joined #code
14:53 gnolam_ [quassel@Nightstar-o3qe20.priv.bahnhof.se] has quit [[NS] Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
15:05 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
15:05 mode/#code [+qo Vornicus Vornicus] by ChanServ
15:12 celmin|sleep is now known as celticminstrel
15:59
<&ToxicFrog>
Woooooo patch accepted
15:59
<&ToxicFrog>
If you use koreader, the next daily should have working calibre wireless sync again.
16:00
<&jerith>
\o/
16:12 mode/#code [+o ErikMesoy] by ChanServ
17:50 Vornlicious [Vorn@Nightstar-lul0ub.sub-174-210-3.myvzw.com] has quit [[NS] Quit: Bye]
17:50 Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
17:52 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds]
21:12
<&McMartin>
I have just been handed the following set of scurrilous rumors and enjoyed them enough to share: http://pythonsweetness.tumblr.com/post/169166980422/the-mysterious-case-of-the-linux-page-table
21:29
<@Tamber>
Well, that is interesting indeed. It'll prove to be even more interesting when the truth comes out.
21:30
<@Tamber>
(I do like a bit of intrigue~)
21:34
<&McMartin>
I think the author's favorite guess is also mine; a way for EC2-like VMs to be able to get access to their cousins by suborning hypervisor code.
21:34
<&McMartin>
That's also about the only threat model that would justify disabling the TLB~
21:34
<&McMartin>
https://i.redditmedia.com/CX-lD7KcNS9R-mnGTTTHi9m3W4LUqVtfXLdY5MDtHfo.jpg?w=600&s=5b0687c5cefaa57d1d8a540b15f0e6cd
21:34
<&McMartin>
("It's because they swapped two and three")
21:35
<@Tamber>
:D
21:35
<&McMartin>
("Since when is there a step 5?" "That's 'by the app store'.")
21:36
<@Tamber>
hee
21:40
<&McMartin>
It is kind of interesting how basically nothing is a "program" anymore.
21:41
<&McMartin>
And thinking about it more, the closest thing you do with computer instructions these days that look like a 'program/programme' in the non-computer sense we've been calling "scripts" since at least the 1970s.
21:41
<&[R]>
How you you reason that?
21:41
<&McMartin>
They're all "apps" or "skills" or "services"
21:41
<@TheWatcher>
"skills", seriously? O.o
21:41
<&McMartin>
The act of writing them is "coding", though that I got to watch happen and accept that this is an outsider community's terminology taking over.
21:41
<&McMartin>
TheWatcher: Those are the modules that Alexa loads, yes.
21:42
<@TheWatcher>
Huh. I guess that makes a kind of sense.
21:43
<&McMartin>
The migration away from the term "program" seems to coincide with the migration away from computing as "the manual allocation and commanding-to-execute of statically linked executables"
21:43
<&McMartin>
And the remaining consumer use case for those always calls them "apps" and if I'm being fair that's an obvious abbreviation of "application program"
21:44
<&McMartin>
I do not deny that the application/service distinction is kind of important
21:44
<&McMartin>
But it also seems like there are really very few cases anymore where a coder actually defines "main" anymore.
21:46 Jessikat` [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
21:46 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Connection closed]
21:47
<&McMartin>
I do generally also accept that user-facing GUI application developers have better things to do than manual poll the event loop~
21:47
<&McMartin>
*manually
21:47
<&McMartin>
Though I do prefer it to be at least theoretically possible, Apple.
21:48 himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds]
21:49
<&[R]>
Any benefit to doing so?
21:49
<&McMartin>
Makes certain kinds of systems easier to bolt GUI support onto as a module.
21:50
<&McMartin>
Whether that's a benefit is a permanent debate
21:50
< Jessikat`>
Like, say, game simulations
21:50
<&McMartin>
(And you can make it happen in Cocoa, at least.)
21:50 Jessikat` is now known as Jessikat
21:50
<&McMartin>
(But you don't do it by actually manually getting and dispatching the events; there's a call you can make that is 'do that main-event-loop stuff until you would block, then return control to me' and that's generally good enough)
21:51
<&McMartin>
It's *most* important when for whatever reason you've decided you want your GUI application to be single-threaded.
21:52
<&McMartin>
A debate for which the two sides may be summarized as "fuck not having background worker threads" and "fuck threads"
21:53
<&McMartin>
I can't speak to the modern game engine philosopy of this stuff, but the VICE emulator wants hardware emulation to be the primary thing the program as a whole is doing, and letting the GUI get around to showing the user what the simulated system looks like is a distant second.
21:53 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Connection closed]
21:53 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
21:53
<&McMartin>
Relying on the GUI's "periodically execute this state update function" routine to keep things like serial ports and music running properly was deemed by the ancients as just asking for unnecessary headaches.
21:54
<@Tamber>
Because it'll work 95% of the time, and the other 5% will either be demos, or out on a client's system~?
21:55
<&McMartin>
Heh. Given that VICE is a Commodore 64 emulator I completely misread "demos" there at first
21:55
<@Tamber>
hehe :)
21:55
<&McMartin>
Note that getting *that* kind of demo to work reliably is an extremely inflexible product requirement~
21:55
<@Tamber>
Quite.
21:56
<@Tamber>
But if it's going to fail, that's when it's going to go. :p
21:56
<&McMartin>
Yep
21:56
<&McMartin>
We've been modernizing the Unix UI and hoping to get it working more broadly
21:56
< Jessikat>
I do enjoy that WPF was built to be agnostic of threads right up until they needed to release it, at which point it suddenly needed to know about the main GUI thread as a concept
21:57
<&McMartin>
GTK3 kind of splits the difference; they *want* you to do the "just let GtkApplication control everything" thing, but if for some reason you can't you can tell it to execute exactly one pending GUI event if possible
21:57
<&McMartin>
And every signal handler you connect gets handed the event structure
21:58
<&McMartin>
A cynic might call that the worst of both worlds, or even suggest that GNOME and its ancillary teams like GTK have been thoughtlessly aping the Macintosh with no knowledge of what the differing design decisions mean
21:59
<&McMartin>
But it mostly does work.
21:59
<&McMartin>
Jessikat: Is WPF here Windows Presentation Framework or some other system?
22:03 Derakon_ is now known as Derakon
22:03 mode/#code [+ao Derakon Derakon] by ChanServ
22:07
<&McMartin>
Tamber: That also said, it occurs to me that most major open-source software lives that 95/5 rule and that's why crash reporters are so aggressively pushed
22:08
<&McMartin>
The flipside of "works on my computer" is "The things that break on *our* systems get fixed in the alpha"
22:16
<@Tamber>
*nods*
22:17
<&[R]>
Which 95/5 rule?
22:26
<&ToxicFrog>
"95% of events come from 5% of causes"
22:27
<&McMartin>
This one was
22:27
<&McMartin>
13:54 <@Tamber> Because it'll work 95% of the time, and the other 5% will either be demos, or out on a client's system~?
22:28
<&[R]>
Ah
22:29
<&McMartin>
For most of my professional career, my employers' clients were companies so large that just the test labs from one of them had hundreds of times as many people working in them as our entire company employed
22:30
<&McMartin>
So the idea that one catches all the bugs in-house was never a world I really lived in outside of programming for absolutely fixed commodity hardware like 80s game consoles.
23:11
<&[R]>
https://imgur.com/gallery/Gu2kMKe
23:15
< Jessikat>
McMartin: yeah, Windows presentation system
23:15
< Jessikat>
The GUI underlying everything since roughly vista I think
23:16
< Jessikat>
There's a marvelously cogent book on its development and internals by one of the chief architects
23:16
<&McMartin>
Oh nice
23:16
< Jessikat>
Which is how I learned it, heh
23:16 himi [sjjf@Nightstar-dm0.2ni.203.150.IP] has joined #code
23:16 mode/#code [+o himi] by ChanServ
23:17 * McMartin has only really used WinForms, which is IHRC one level of abstraction down, and then raw user32.dll calls, which is an additional layer below that.
23:17
<&McMartin>
That latter was pretty great though, given that it resulted in a binary that ran unchanged and without compat-mode flags on both Windows 10 and Windows 98
23:19
< Jessikat>
WPF is genuinely an excitingly powerful system once you understand the intricacies though
23:20
< Jessikat>
It's what we use a lot in tools at work and it lets us get a lot of the mundane stuff done super quick
23:20
<&McMartin>
I can imagine; my memory of it when I was looking at it was "this is doing a lot of great stuff that I actually don't need for the thing I need to write right now"
23:20
<&McMartin>
It seemed focused on, hrm
23:21
<&McMartin>
"Document based applications" isn't *quite* right - but something where there's some core data structure altered by both the user and internal logic and you wanted seamless view updates in both directions
23:21
< Jessikat>
It's built around separating data from presentation, aggressively so
23:21
< Jessikat>
And the visual tree is flexible that basically anything can contain anything else
23:22
< Jessikat>
There's a couple of quirks in calculating boundaries and sizes you need to know
23:22
< Jessikat>
And the controls are aggrivatingly nonstandard in places
23:22
< Jessikat>
But in general it's a very nice system
23:23
< Jessikat>
Certainly the least unhappy I've been when developing GUI apps
23:23
< Jessikat>
And I find basically any GUI tedious beyond belief
23:24
< Jessikat>
I enjoy coding to hardware and designing applications but very little of what's in the middle for its own sake, heh
23:35 Degi_ [Degi@Nightstar-6rc6ah.dyn.telefonica.de] has joined #code
23:38 Degi [Degi@Nightstar-2aa9mg.dyn.telefonica.de] has quit [Ping timeout: 121 seconds]
--- Log closed Tue Jan 02 00:00:20 2018
code logs -> 2018 -> Mon, 01 Jan 2018< code.20171231.log - code.20180102.log >

[ Latest log file ]